Issue with checking epub mimetype

[prernagp]

Some epub files are not extracting mime-type properly. Getting 'application/epub+zip\r\n' when trying to check for epub files.

Some of the links for the epub files are here:-
https://soundplay-reports.s3.amazonaws.com/1.epub
https://soundplay-reports.s3.amazonaws.com/2.epub
https://soundplay-reports.s3.amazonaws.com/3.epub

The extra characters should be truncated by the library.

[prernagp]

@sindresorhus Please check this issue

[sindresorhus]

You can see in the source that there's no \r\n being added, so I don't think the issue is with file-type:

file-type/core.js

Line 340 in 330e6b9

mime: 'application/epub+zip',

I'm happy to reopen if you submit a failing test.

[prernagp]

@sindresorhus It was the issue with token-types library. We have raised the issue there with the solution

Borewit/token-types#455

[Borewit]

@prernagp is talking about this part, where a decoding issue is (caused by the extra \n\r in mimeType):

file-type/core.js

Lines 334 to 336 in cffcf1f

	const mimeType = await tokenizer.readToken(new Token.StringType(zipHeader.compressedSize, 'utf-8'));
	switch (mimeType) {

That results in a detection of application/zip instead the expected application/epub+zip.

[prernagp]

Thank you @Borewit for the help